Major campaign underway to nullify Electoral College

[Skill and Chance]

I think this underscores the real motivation for legislative action - which party stands to gain by NPVIC. The GOP had a clear structural advantage in the EC for a long time such that no non-Southern Dem won the presidency after Kennedy until Obama.

That's not evidence of a structural advantage: counting Gore as a southern Democrat, no non-southern Dem won the popular vote after Kennedy until Obama. No non-right-handed Republican won the presidency after Nixon until Bush Jr. - it's a structural advantage! Or, maybe, coincidence, or the way these things work.

You reversed my two clauses and their dependency. The structural advantage is measured by how the EC would go if the vote shifted to an even split between the two candidates. Until the last decade that advantage was typically for the GOP and an even race would be expected to go for the GOP. Since a big piece of that base since 1960 was in the South, a southern candidate could swing regional votes and get a win like Carter did in 1976. Gore was not really considered a Southern candidate after his 8 years as VP.

No, there wasn't. From 1960 to 2004, Republicans had the theoretical EC edge six times (1968, 1976, 1984, 1988, 1992 and 2000) while Democrats had it the other six times (1960, 1964, 1972, 1980, 1996, 2004). There appears to be no consistent pattern in the EC advantage.

Indeed. FTR, here's the structural advantage (measured as the difference in winning margins between the national popular vote and the popular vote in the decisive State) from 1932 to 2012:

1932: R+0.05
1936: R+3.69
1940: R+3.06
1944: R+2.48
1948: R+3.64
1952: R+0.62
1956: D+0.86
1960: D+0.64
1964: D+2.01
1968: R+1.58
1972: D+0.88
1976: R+0.38
1980: D+1.81
1984: R+0.77
1988: R+0.17
1992: R+0.91
1996: D+0.68
2000: R+0.53
2004: D+0.35
2008: D+1.98
2012: D+1.52

No clear pattern here, except maybe that Republicans had a structural advantage in the years of the Solid South, as Democrats wasted a sh*tload of votes there (but this would require going farther back to confirm).

It's fascinating that 2000 and 1960 (PV winner disputed based on allocation in AL) actually were not years with large biases.  A serious 3rd party candidate also seems to result in additional bias.

[Skill and Chance]

What Antonio's fine work suggests to me is that from 1964-80 there was no clear advantage to either party in the EC. After the reapportionment of 1980, there seems to be a Pub advantage through 2000 with 4 of 5 elections having an R lean. After 2000 the EC seems to shift to the Dems, so it will be interesting to see if that continues for the remaining elections this decade.

More data is needed, but one could argue that a Solid South effect is now developing for the GOP.